Waste Glass Reuse in Foamed Alkali-Activated Binders Production: Technical and Environmental Assessment

Ruan, Shaoqin; Kastiukas, Gediminas; Liang, Shuang; Zhou, Xiangming

doi:10.3389/fmats.2020.581358

ORIGINAL RESEARCH article

Front. Mater., 28 September 2020
Sec. Structural Materials
Volume 7 - 2020 | https://doi.org/10.3389/fmats.2020.581358

Waste Glass Reuse in Foamed Alkali-Activated Binders Production: Technical and Environmental Assessment

Shaoqin Ruan^1,2

Gediminas Kastiukas²^*

Shuang Liang²

Xiangming Zhou²^*

¹College of Civil Engineering and Architecture, Zhejiang University, Hangzhou, China
²Department of Civil and Environmental Engineering, Brunel University London, Uxbridge, United Kingdom

Waste glass is a type of construction and demolition waste, which carries significant environmental burdens and can be recycled. In this study, a foamed alkali-activated glass (FAAG) via the use of milled waste glass was synthesized in low-temperature (i.e., 80°C). The influence of foaming agent (i.e., aluminum powder) content on the foaming process of samples was investigated and optimized in view of their performance. The optimum mix formulation, which is FAAG with 10% aluminum powder was then selected, and followed by the measurement of the water-resistance and absorption, impact resistance and thermal conductivity of the prepared FAAG. X-ray diffraction was also involved to identify the phase formation within the samples. Finally, the environmental impacts related to the preparation of 1 kg FAAG were obtained. Results shows that a type of more sustainable FAAG compared with the commercial product was successfully developed in low-temperature, revealing the low thermal conductivity of ∼0.13 W/(mK) with a density of ∼0.59 g/cm³ and compressive strength of ∼5.52 MPa, which is 97% higher than the commercial product. Meanwhile, in addition to a foaming agent, aluminum powder is also acting as a key component in the geopolymerization of samples, facilitating the formation of aluminosilicate that provides strength.

Introduction

In the European Union, construction and demolition waste (CDW) is becoming a serious issue, accounting for around 25–30% of all waste generated. CDW can be categorized into several groups such as concrete, bricks, wood or glass. Among them, in 2018, 27 metric million tons of glass is produced and only 21% of them are recycled (Statista, 2018). At present, CDW including waste glass was mainly disposed and treated via the procedures of landfilling or incineration, and (Zaman, 2010) reported that landfilling presented significantly lower environmental burdens than incineration due to fewer emissions produced and no direct heat required. In order to increase add-value of CDW and potentiate the circular economy, the reuse of secondary resources is highly recommended by European Commission targets (European Commission, 2020). For instance, 67% of waste glass was recycled in United Kingdom (Lebullenger and Mear, 2019). However, it is also indicated that owing to a large land requirement and difficulty in monitoring emissions in a long time, there are more restrictions in terms of landfilling of waste in the developed countries (Zaman, 2010). Therefore, the reuse of waste glass in the area of construction provides an alternative solution to its disposal. Currently, a simple option of waste glass under exploration is in the replacement of virgin aggregates in concrete (Lu et al., 2019; Xiao et al., 2020). Besides, the use of waste glass treatment in the construction sector is also investigated. For instance, it can be used in the preparation of alkali-activated binders or geopolymers.

Alkali-activated binders or geopolymers have been widely studied due to their potential applications which can be environmental friendly (Ma et al., 2018) and can replace Portland cement (PC) in several niche applications (Panda et al., 2019; Panda et al., 2020), and waste glass has also been utilized for their preparation. For instance, waste glass is capable of acting as an alkaline activator, which was proved to be effective for the hydration of the precursors (Torres-Carrasco and Puertas, 2015; Tchakouté et al., 2016). At the same time, waste glass can also be regarded as a precursor, however, its reactive alumina content is relatively low [i.e., 0–5% (Metro)], which inhibits its geopolymerization. Therefore, some additives containing alumina should be added such as calcium aluminum cement, which resulted in the formation of sodium-aluminosilicate (N-A-S-H) gel as well as a densified microstructure, contributing to the improvement of the compressive strength of samples (Luhar et al., 2019). At the same time, (Vafaei and Allahverdi, 2017), also synthesized geopolymers via the usage of waste glass and calcium aluminate cement, and a more cross-linked network was formed within the samples with higher compressive strength revealed. In addition to calcium aluminum cement, waste glass is also widely mixed with metakaolin to produce geopolymers. For instance, Si et al. (2020) demonstrated that the incorporation of waste glass into the metakaolin-based geopolymer samples led to the formation of a denser gel structure with a refinement of pore structure. Besides, the shrinkage of geopolymers, which is a serious issue in hardened samples (Yang et al., 2017; Zhu et al., 2018), also reduced due to the refinement of their pore structure so that a reduction in the moisture loss rate was achieved within the samples (Si et al., 2020). Also, Bai et al. (2019) developed waste glass-based cellular geopolymers with the use of a stabilizing agent, which revealed thermal conductivity of ∼0.21 W/mK and compression strength of ∼7.3 MPa.

In the meantime, some precursors such as tungsten mining waste are also short of a silica source during the geopolymerization, therefore, Kastiukas and Zhou (2017) reported that the incorporation of waste glass into alkali-activated tungsten mining waste was capable of providing an additional silica source, leading to the formation of (C, N)–A–S–H phases within the samples as well as a boost of their mechanical strength. Further, accompanied by an increment of strength, it was also revealed that the addition of glass cullet and powder into alkali-activated binders resulted in an improvement of their workability and fire-resistance (Lu and Poon, 2018) with low thermal conductivity presented (Xuan et al., 2019).

Foamed geopolymers or alkali-activated binders, which is a recent innovation, present the benefits of foamed concrete and geopolymers, and they are capable of reducing the environmental impacts of construction materials and offering new alternative materials for thermal insulation (Gu et al., 2020). Due to the excellent thermal insulation, foamed geopolymers can be prepared through a physical or a chemical foaming process. The physical foaming route consists of a mixture of geopolymers and precast foam prepared by a foaming agent, which would then produce a large number of pores inside the paste after setting, whereas the chemical one mainly mixes chemical foaming agent such as H₂O₂ and aluminum powders, followed by the production of H₂ and O₂ that introduced pores within the paste, as shown in Eqs. 1 and 2.

{2 Al + 6 H}_{2} O + 2 NaOH \to {2 NaAl (OH)}_{4} {+ 3 H}_{2} ↑ (1)

H_{2} O_{2} \to H_{2} {O + O}_{2} ↑ (2)

Unlike foamed geopolymer, there are also numerous studies investigating the performance of foamed glass as it has already been used as a thermal insulation material for the buildings, however, its preparation is an energy-intensive process where high temperature (i.e., 700–800°C) is required and emits CO₂, which contributes to the global climate change. During the preparation of foamed glass, the raw glass is first ground and mixed with the carbon and Mn₃O₄ (i.e., foaming agent), and the mixture is then heated in high temperature with some protective atmosphere introduced such as non-oxidizing gases (e.g., Ar or N₂). During the heating process, CO₂ is produced to generate some pores within the glass, as shown in Eq. 3. Finally, it is necessary to cool down the newly produced foamed glass (König et al., 2019; Østergaard et al., 2019). Therefore, a foamed geopolymer based on the use of glass may be a better option than a foamed glass in terms the economic and environmental concerns.

{C + Mn}_{3} O_{4} \to {CO}_{2} + MnO (3)

As aforementioned, although waste glass is already used as a type of alkali-activator, or a partial replacement of slag, fly ash, etc, only a few studies investigated the use of 100% waste glass into the synthesis of alkali-activated binders especially foamed products. Besides, the produced foamed alkali-activated glass (FAAG) is different compared with the production process of foamed glass in terms of the production process, which avoids high temperature and CO₂ emission. Further, a comprehensive life cycle assessment in terms of the preparation is scarce in the literature, particularly considering the environmental impacts of waste glass treatment during the assessment.

Therefore, in view of the foregoing, this study fabricated a foamed alkali-activated binder based on 100% of waste glass, and the parameter that was investigated in this study was the content of the foaming agent (i.e., aluminum powder). The FAAG with 10, 20 and 30% aluminum powder were tested in this study. Consequently, the optimal content of foaming agent was selected and the corresponding samples were then prepared based on their mechanical performance and density. After that, the key performance of samples such as thermal conductivity, water, and impact resistance were explored and the outcomes were compared with the existing foamed products in the literature. Further, this study also performed a comprehensive life cycle assessment of FAAG preparation, where the environmental burdens related to the landfilling or incineration of waste glass were also taken into account, and the outcomes were then compared with a commercial aerated product.

Mix Design, Materials, and Methods

Raw Materials and Mix Design

The raw materials used to produce the FAAG in this study consisted of waste glass (received from the local municipality of Covilhã, Portugal), sodium hydroxide (NaOH, 98%) (Fisher Scientific, Germany), sodium silicate (Na₂SiO₃) (Solvay SA, Portugal), and aluminum powder (Sigma Aldrich, United Kingdom). The 10 M sodium hydroxide solution, prepared by dissolving sodium hydroxide pellets in de-ionized water, was combined with the sodium silicate solution consisting of a SiO₂/Na₂O ratio of 3.23 (8.60 wt% Na₂O, 27.79 wt% SiO₂, 63.19wt% H₂O, and 0.4wt% Al₂O₃), to form the alkali activator. The wavelength dispersive X-ray fluorescence was used to obtain the chemical composition of the waste glass (Na₂O:12.4%; MgO:1.8%; Al₂O₃:2.1%; SiO₂: 68.7%; SO₃: 0.3%; K₂O: 0.8%; Fe₂O₃: 1.5%; CaO: 10.0%). Figure 1 shows the waste glass as received and after milling, respectively.

FIGURE 1

FIGURE 1. Waste glass as received (A) and after milling (B).

The waste glass was milled at 500 rpm for 20 min using a planetary ball milling machine (Retsch, PM 100, Germany) to achieve a powder consistency. The particle size distribution curve is illustrated in Figure 2. The activator solution was added to the precursor (i.e., waste glass) and blended using a bench-top mixer at 300 rpm for 5 min, forming the paste. Immediately after, the aluminum powder was added to the paste and followed by a further 30 sec of mixing. The paste was cast into polystyrene molds with the dimensions of 40 mm × 40 mm × 160 mm allowing for the foaming reaction between aluminum powder, water, and NaOH to take place, as expressed in Eq. 1 already.

FIGURE 2

FIGURE 2. Particle size distribution of milled waste glass powder.

The freshly foamed samples were sealed and placed in a temperature and humidity controlled environmental chamber at 80°C and 75% relative humidity for 24 h, followed by 20°C and 75% relative humidity until the specific age (i.e., 7, 14, and 28 days) of testing. The Na₂SiO₃/NaOH, waste glass-to-activator and water-to-waste glass ratios were fixed at 3, 4.5, and 0.12 separately. Three levels of aluminum powder (i.e., 10, 20, and 30%) were investigated, and its optimal content was selected based on the mechanical performance and density. The mix design of FAAG samples in this study is shown in Table 1. The samples were then prepared, and the measurements of their water-resistance and absorption, impact resistance, thermal conductivity, and environmental impacts were performed later.

TABLE 1

TABLE 1. The mix design of foamed alkali-activated glass samples.

Experimental Work

In the experimental work, the mechanical properties, density, water resistance, water absorption, impact resistance, and thermal conductivity of samples were tested. The use of X-ray diffraction (XRD) facilitates the understanding of phases within the samples. The details are shown as follows.

Mechanical Properties

The compressive and three-point bending of each sample were tested after 7, 14, and 28 days as specified by EN196-1 (2016) using a 25 kN universal testing machine (Instron 5960, United Kingdom) at a constant loading rate of 3 kN/min. The dimensions of samples are 40 mm × 40 mm × 160 mm during the mechanical strength test. The compressive and three-point bending values were the average ones obtained from three samples.

Density

The true density of the FAAG in triplicates was determined with a helium pycnometer (AccuPyc 1,340, Micromeritics, United Kingdom) in line with ASTM (2017). The device adopted some approaches of gas displacement and the relationship between the volume and the pressure relationship is known as Boyle’s Law. The pycnometer is identified as a type of density measuring device, however, they are merely measuring volume actually, and density is obtained via a ratio of mass to volume; and the mass is also measured through the device by weighing.

Water Resistance

Six samples at the curing age of 14 days were dried for 24 h at a constant temperature of 80°C, weighed and then placed in a dry chamber for cooling without moisture absorption. The prisms were subsequently placed inside a container filled with de-ionized water and left at 20°C and 75% relative humidity to cure for a further 7 or 14 days. The mechanical performances of samples were determined, along with the strength retention coefficient as defined by Eq. 4 (Deng, 2003):

W_{n} = R_{c n} / R_{c} (4)

Where W_n is strength retention coefficient, in which n is the days of sample immersing water; R_cn is the strength of the wet samples after immersion in water for n days; R_c is the strength of the dry samples cured for different days (e.g., 7 and 14 days).

Water Absorption

To determine the water absorption of the FAAG, three similar prisms were oven-dried at a temperature of 80°C for 24 h, and the dry weight was recorded as the initial weight. The samples were then immersed in water for 48 h, after which the saturated surface dry weight was recorded as the final weight. The calculation is shown below, according to ASTM (2012) in Eq. 5:

W % = [(M_{s a t u r a t e d} - M_{d r y}) / M_{d r y}] \times 100 % (5)

Impact Resistance

The impact resistance of FAAG was determined using the Charpy testing method, in line with ASTM (2018). The Charpy testing method is based on the energy measurement expressed in energy (Joule) required to breaking the samples, shown in Figure 3. Since the maximum potential energy of the pendulum is known can be expressed as a function of the mass of the hammer and fall height, the energy absorbed can be calculated via the breakage of the samples (CEAST, 2008).

FIGURE 3

FIGURE 3. (A) Setup of Charpy testing (CEAST Torino, United Kingdom) (B), (C) view of samples cut out to an appropriate size from different angles.

Thermal Conductivity

The thermal conductivity was measured with a thermal conductivity meter (Fox 200, TA Instruments, United States). The steady-state heat flux through the rectangular block sample was measured for a temperature gradient of 10°C between the upper and the lower face of the sample. Three identical 150 mm × 150 mm × 10 mm samples for each sample were measured for a more accurate result of thermal conductivity. Before measurement, the samples were left 24 h at 80°C and placed for 30 min in a dry chamber for cooling without moisture absorption.

X-ray Diffraction

To identify the newly formed phases in FAAG, a Bruker X-ray diffractometer (D8 Advance, Germany) was used with an automatic slit, monochromated Cu Kαα radiation (λ = 1.5405 Å), 15°–50° 2θ range, with steps of 0.02°/2θ and 0.5 sec/step. The raw material (i.e., waste glass) and the crushed FAAG samples after compression test with different amounts of the foaming agent were ground and sieved, and the powder was stored in ethanol for 3 days to stop hydration and dried prior to testing. Peaks were identified using the software DIFFRACT.SUITE (Bruker, Germany).

Environmental Assessment

SimaPro 8, as a software tool is used for life cycle assessment in this study. This section introduces the goal and system boundary, inventories, transportation, and impact assessment during the life cycle assessment (from cradle to gate) of the FAAG preparation as follows:

Goal and System Boundary

The goal of this part was to perform a life cycle assessment for the preparation of 1 kg FAAG, where the damage categories, CO₂ emissions and energy consumptions associated with its preparation were investigated, and the outcomes were compared with the equivalent amount of aerated commercial products (sand/gravel, quicklime, anhydrite, cement, and aluminium powder). Meanwhile, the impacts of waste glass disposal were avoided from the outcomes related to FAAG preparation during the assessment, and a similar idea was also widely reported in a number of environmental assessments (Zhao et al., 2009; He et al., 2019). Figure 4 displays the system boundary of environmental assessment due to the preparation of 1 kg FAAG considering impacts of waste glass disposal and 1 kg commercial aerated product.

FIGURE 4

FIGURE 4. System boundary of environmental assessment in view of preparation of 1 kg foamed alkali-activated glass considering impacts of waste glass disposal as well as 1 kg commercial aerated products.

Inventories

Since the glass used in this study was regarded as pure municipal waste, any inputs and outputs associated with its production can be neglected. The inventories of the aerated commercial product were obtained from the database Ecoinvent (Version 2.0). To produce 1 kg of FAAG, the inputs and outputs related to the production of the rest of the raw materials, that is, NaOH, NaSi₂O₃, and aluminum powder, as well as electricity and transportation, were gained from the database Ecoinvent and ETH-ESU. To produce 1 kg of FAAG, 1 kg of waste glass, 119.6 g of water, 5.6 g of aluminum powder, 55.4 g of NaOH, and 166.4 g of Na₂SiO₃. The electricity is calculated to be 0.25 kWh, which is calculated through the power (i.e., 750 w) *time (i.e., 20 min) used in the milling of waste glass.

Transportation

Only the transportation of waste glass is considered, and the distance is assumed to be 2,000 km from Covilhã, Portugal to London, United Kingdom, and the travel distances of the rest raw materials, i.e., NaOH, NaSi₂O₃, and aluminum powder from a local factory or distributor to London, United Kingdom were assumed to be 650, 2,000, and 180 km, respectively. In view of the landfilling or incineration of waste glass, the travel distance of waste glass from Covilhã, Portugal to the nearest landfilling or incineration site was assumed to be 300 km. The means of transportation concerning all the materials were assumed to be by lorry.

Impact Assessment

The damage categories, including damage to health, ecosystem quality, and resources, were obtained by the Eco-indicator 99 approach. Furthermore, instead of a single score, a mixing triangle was used to tackle the weighting issues when making the decisions in terms of the selection of FAAG or the commercial aerated products, and the determination of points was obtained when they cross the boundaries of the triangle, and the indifference lines were drawn later (Hofstetter et al., 1999). The detained points calculated on the corresponding triangle sides were acquired via the calculation of Eqs 6–8.

W R = {D_{H H} (A) / T_{H H} - D_{H H} (B) / T_{H H}} / {D_{R} (B) / T_{R} - D_{R} (A) / T_{R} + D_{H H} (A) / T_{H H} - D_{H H} (B) / T_{H H}} (6)

W_{HH} = {D_{R} (B) {/T}_{R} - D_{R} (A) {/T}_{R}} / {D_{R} (B) {/T}_{R} - D_{R} (A) {/T}_{R} {+ D}_{HH} (A) {/T}_{HH} - D_{HH} (B) {/T}_{HH}} (7)

W_{EQ} = {D_{R} (B) {/T}_{R} - D_{R} (A) {/T}_{R}} / {D_{R} (B) {/T}_{R} - D_{R} (A) {/T}_{R} {+ D}_{EQ} (A) {/T}_{EQ} - D_{EQ} (B) {/T}_{EQ}} (8)

Where D_R(A), D_EQ(A), D_HH(A), D_R(B), D_EQ(B), and D_HH(B) represent the damages of product A (i.e., FAAG) and B (i.e., the commercial aerated product) in terms of resources (R), ecosystem quality (EQ), and human health (HH), separately. The overall damages of each safeguard theme are referred to as T_R, T_EQ, and T_HH; and W_R, W_EQ, and W_HH represent the weighting of the safeguard themes on the corresponding triangle side.

The CO₂ emissions of the preparation of 1 kg FAAG and 1 kg commercial aerated product were obtained via an Intergovernmental Panel on Climate Change 2007 approach (Pachauri and Reisinger, 2007) with a timeframe of 100 years. The energy consumptions associated with the sample preparations were calculated via the Cumulative Energy Demand method. The Cumulative Energy Demand method incorporates the sum of direct and indirect energy consumption (mega joule) throughout the entire life cycle, including the energy consumed during the production of raw materials and transportation (Bribián et al., 2011).

Results and Discussion

Selection of Foamed Alkali-Activated Glass Formulation

The effects of aluminum powder content on the mechanical performance and the density of FAAG were investigated in this section, facilitating the selection of FAAG formulation. Figure 5 presents the pore structures of selected samples, indicating the excellent foaming performance of FAAG, which consists of a closed network of pores, regular in shape and distribution depending on the content of aluminium powder. Figure 6 shows the density of FAAG reduced with the growth of aluminum powder content after 7 days of curing. The density of FAAG with 30% aluminum powder (i.e., 0.48 g/cm³) is 17.8% lower than the counterparts with 10% (i.e., 0.59 g/cm³). The latter is related to a higher amount of H₂ gas formed within this mixture, presenting a more porous structure, as seen in Eq. 1.

FIGURE 5

FIGURE 5. Foamed alkali-activated glass containing (A) 10%; (B) 20% and (C) 30% aluminum powder via the use of the optical microscope, and the images were captured by a software (Scale label: x210).

FIGURE 6

FIGURE 6. Effect of aluminum powder content on the mechanical performance and density of foamed alkali-activated glass cured for 7 days.

As expected, Figure 6 also illustrates that FAAG with decreasing density led to inferior mechanical performance including compressive and flexural strength, and both compressive and flexural strength of FAAG with 30% aluminum powder is only 1/3 than the group with 10%. It is to be noted that when the amount of aluminum powder increased from 10 to 20%, there is nearly no decline in terms of the strength, and a possible explanation could be related to the instability of foams formed within the samples, which caused by the foam bubbles and a sudden reduction in vertical thickness of the system, as indicated by the large error bars in Figure 6, and Hajimohammadi et al. (2018) also reported that the foamed geopolymers via the chemical approach were not as controllable as the physical foamed ones. However, a growth of aluminum powder from 20 to 30% is accompanied by a great reduction in the strength and density due to the porous structure formed resulting from H₂ generation. Besides, the difference of irregularity of pore morphology and interconnections in samples containing 30% may also account for its great decline in the mechanical performance.

Albeit it presents the lowest density, FAAG with 30% aluminum powder was not considered due to its relatively low strength and a high amount of aluminum powder used, that is, a higher overall cost in terms of the sample preparation. The aluminum powder content in later FAAG preparation was fixed to 10% based on its similar mechanical performance with FAAG containing 20% of foaming agent, as well as cost-effectiveness via reducing the amount of foaming agent introduced. At the same time, the density of FAAG with 10% of foaming agent (i.e., 0.59 g/cm³) was not sacrificed compared with the group with 20% (i.e., 0.57 g/cm³), which is highly favored for the thermal insulation.

Performance of Selected Foamed Alkali-Activated Glass Samples

In this section, the results of mechanical performance, water resistance, water absorption, impact resistance, and thermal conductivity of the prepared FAAG were elaborated from Mechanical Properties to Thermal Conductivity.